Credit: CC0 Public Domain

Örebro researchers Rajesh Patil and Professor Magnus Löfstrand have developed an AI system that detects welding defects, reduces material waste, and supports sustainable manufacturing.

"We’re now looking to test the system together with the automotive industry," says Rajesh Patil.

Patil and Professor Löfstrand envision manufacturing plants where machines, without human involvement, rapidly identify welding defects, remove faulty components, and minimize material waste. Their research is conducted in mechanical engineering at Örebro University’s School of Science and Technology.

Together, they have developed an AI-driven inspection system that detects and classifies weld defects in engine exhaust components with high speed and precision. By combining artificial neural networks (ANNs) with support vector machines (SVMs), the system can identify defects in both similar and dissimilar metals—tasks that would be both time-consuming and extremely difficult for human inspectors.

"Our AI acts as smart eyes on the factory floor, identifying welding defects in real time, improving product quality, reducing waste, and enabling faster, more sustainable production," says Löfstrand.

The study, which is being published in the journal Journal of Failure Analysis and Prevention, builds on previous research in AI-based weld inspection and has resulted in a fully autonomous system adaptable to various industrial welding applications. The method can help reduce both energy consumption and material waste, aligning with Sweden’s goals for sustainable manufacturing.

"This is a major step toward smart factories where AI works side by side with people to improve efficiency and sustainability," says Patil.

The research group is now seeking collaboration with local and international automotive manufacturers to test the system in industrial production environments.

More information: Rajesh Patil et al, Detection and Classification of Engine Exhaust Weld Joint Defects Using RNN and SVM on SS316L–SS410 and SS310–SS410, Journal of Failure Analysis and Prevention (2025). DOI: 10.1007/s11668-025-02292-7

Citation: AI system with smart eyes detects welding defects (2026, January 20) retrieved 20 January 2026 from https://techxplore.com/news/2026-01-ai-smart-eyes-welding-defects.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.